|
Dear {name},
We are happy to say hello again. This is our third humanet3 newsletter, and we are dedicating this one to the art of asking questions. With the arrival of GPTs, our daily lives are filled with questions in the form of prompts—there seems to be a cultural shift not only in where we are looking for answers, but also in how we are looking for them. An old piece of wisdom says that the only stupid question is the one you don't ask. At the same time, however, the philosopher of science Thomas Kuhn reminds us that the answers you get depend upon the questions you ask. A few days ago, I read the latest piece by Stefaan Verhulst (The GovLab): Inquiry as Infrastructure: Defining Good Questions in the Age of Data and AI. In summary:
He suggests that literacy is a fundamental but underdeveloped skill in the data and policy landscape, and advocates for a ‘new science of questions’ to delve deeper into what constitutes a good question. In his view, they must be technically feasible, ethically grounded, socially legitimate, and aligned with real-world needs. A proposal that raises tons of further questions and connects with a topic we have been discussing lately in our group—infrastructures: see e.g., data is infrastructure (Bietti) and digital public infrastructure or DPI (Mazzucato et al.)
The DPI topic has sprouted along new EU policy debates after the Draghi Report highlighted geopolitical questions around the EU’s ability to reduce its technological dependence on US and China. In our February workshop on Re-imagining Digital Public Spaces for Democracy (concept note, pictures, and video statements by the panelists here, and a very cool group video here), we had an intense debate on the matter: is it desirable for the public sector to be involved in the development of new technologies, notably AI? Is it rather advisable to promote EU champion firms in the digital sector? Has the new EU regulatory landscape stifled incentives to innovate in the sector? The narrative of regulatory simplification is gaining ground and, for many voices, this is rather a covert way of saying: let's deregulate. Would you agree with this diagnosis?
We are certain that some of these debates will be addressed as research questions in the special journal issue that we just arranged to publish in partnership with Regulation & Governance, as a follow-up to our last workshop. We will keep you posted!
Related policy questions that have caught our attention refer to the post-DSA wave of voices envisioning a more decentralized social media ecosystem. These range from platform design (e.g., Weyl et al.'s Prosocial Media) to the question of how DMA’s interoperability rules for messaging services may also help social media decentralisation in the EU (e.g. Scott Morton). In 2025 we cannot neglect AI agents in discussions about content curation. Yet, the question whether they will curate content in alignment with protected values, leads us back to other crucial ones: Who controls the technology? And how effective can AI regulatory efforts be? Three authors we have read closely —Deb Roy, Larry Lessig and Audrey Tang— have recently suggested interesting pathways in Conversation Networks. Beyond Clicks and Comments: Leveraging AI for Meaningful Civic Engagement. We will be discussing them in our next humanet3 journal club—just send us an email if you want to join us at our premises in Berlin!
In the coming months, we will continue exploring digitised relationships —both between humans and between humans and machines— and how, against this backdrop, we can preserve values such as community, autonomy, and free speech.
One final personal note on this: while protecting people from potential Big Tech abuses and censorship is essential, this cannot be achieved without also acknowledging censorship and silencing of voices by states—particularly in EU Member States like Germany, where respectful voices raised for Palestinians have led to entry bans and deportations. Whatever your view, the key point is that without free speech, we cannot share our most fundamental questions. Questions that require human-to-human dialogue, as they are not (and may never be) meaningfully answerable by a machine.
Best wishes,
Germán |
